Owing to the increasing scarcity of non-renewable energy reserves such as coal, petroleum and uranium, increased use is being made of alternative nondepletable energy sources, such as photovolatic energy. Single crystal photovoltaic devices, especially crystalline silicon photovoltaic devices, have been utilized for some time as sources of electrical power because they are inherently non-polluting, silent and consume no expendable natural resources in their operation. However, the utility of such devices has been limited by problems associated with the manufacture thereof. More particularly, single crystal semiconductor alloy materials (1) are difficult to produce in sizes substantially larger than several inches in diameter; (2) are thicker and heavier than their thin film counterparts; and (3) are expensive and time consuming to fabricate.
Recently, considerable effort has been expended to develop systems and processes for preparing thin film amorphous semiconductor alloy materials which encompass relatively large areas and which can be deposited so as to form p-type and n-type semiconductor alloy layers for the production therefrom of thin film electronic devices, particularly thin film p-n type and n-i-p type photovoltaic devices which are substantially operatively equivalent or superior to their crystalline counterparts. It should be noted at this point that the term "amorphous" as used herein, is defined to include alloys or materials exhibiting long range disorder, although said alloys or materials may exhibit short or intermediate range order or even contain crystalline inclusions.
Amorphous thin film semiconductor alloys have gained acceptance as the material from which to fabricate electronic devices such as photovoltaic cells, photoresponsive and photoconductive devices, transistors, diodes, integrated circuits, memory arrays and the like. This is because the amorphous thin film semiconductor alloys (1) can be manufactured by relatively low cost continuous processes, (2) possess a wide range of controllable electrical, optical and structural properties and (3) can be deposited to cover relatively large areas. Among the semiconductor alloy materials exhibiting the greatest present commercial significance are amorphous silicon, germanium and silicon-germanium based alloys. Such alloys have been the subject of a continuing development effort on the part of the assignee of the instant invention, said alloys being investigated and utilized as possible candidates from which to fabricate a wide range of semiconductor, electronic and photoresponsive devices.
Additionally, said assignee has developed commercial processes for the continuous roll-to-roll manufacture of large area photovoltaic devices. Such continuous processing systems are disclosed in the following U.S. Patents, disclosures of which are incorporated herein by reference: U.S. Pat. No. 4,400,409, for A Method Of Making P-Doped Silicon Films And Devices Made Therefrom; U.S. Pat. No. 4,410,588, for Continuous Amorphous Solar Cell Production Systems; and U.S. Pat. No. 4,438,723, for Multiple Chamber Deposition and Isolation System And Method. As disclosed in these patents a web of substrate material may be continuously advanced through a succession of interconnected, environmentally protected deposition chambers, wherein each chamber is dedicated to the deposition of a specific semiconductor alloy material onto the web or onto a previously deposited layer. In making a photovoltaic device, for instance, of n-i-p type configurations, the first chamber is dedicated for the deposition of an n-type semiconductor alloy material, the second chamber is dedicated for the deposition of a substantially intrinsic amorphous semiconductor alloy material, and the third chamber is dedicated for a deposition of a p-type semiconductor alloy material. The layers of semiconductor alloy material thus deposited in the vacuum envelope of the deposition apparatus may be utilized to form photoreponsive devices, such as, but not limited to, photovoltaic devices which include one or more n-i-p type cells. By making multiple passes through the succession of deposition chambers, or by providing an additional array of deposition chambers, multiple stacked cells of various configurations may be obtained. Note, that as used herein the term "n-i-p type" will refer to any sequence of n and p or n, i and p semiconductor alloy layers operatively disposed and successively deposited to form a photoactive region wherein charge carriers are produced by the absorption of photons from incident radiation.
The concept of utilizing multiple stacked cells, to enhance photovoltaic device efficiency, was described at least as early as 1955 by E.D. Jackson in U.S. Pat. No. 2,949,498 issued Aug. 16, 1960. The multiple cell structures therein discussed were limited to the utilization of p-n junctions formed by single crystalline semiconductor devices. Essentially the concept employed different band gap devices to more efficiently collect various portions of the solar spectrum and to increase open circuit voltage (Voc). The tandem cell device (by definition) has two or more cells with the light directed serially through each cell. In the first cell, a large band gap material absorbs only the short wavelength light, while in subsequent cells, smaller band gap materials absorb the longer wavelengths of light which pass through the first cell. By substantially matching the generated currents from each cell, the overall open circuit voltage is the sum of the open circuit voltage of each cell, while the short circuit current thereof remains substantially constant. Such tandem cell structures can be economically fabricated in large areas by employing thin film amorphous, semiconductor alloy materials (with or without crystalline inclusions), in accordance with the principles of the instant invention. It should be noted that Jackson employed crystalline semiconductor materials for the fabrication of his stacked cell structure; however, since it is virtually impossible to match lattice constants of differing crystalline materials, it is not possible to fabricate such crystalline tandem cell structures in a commercially feasible manner. In contrast thereto, and as the assignee of the instant invention has shown, such tandem cell structures are not only possible, but can be economically fabricated over large areas by employing the amorphous semiconductor alloy materials and deposition techniques described herein.
As described in the previously referenced patents and applications, Applicants' assignee is now able to manufacture large area stacked, photovoltaic devices on a commercial basis utilizing a roll-to-roll processor. In particular, said assignee has developed a high volume combination r.f. and microwave energized apparatus for the fabrication of photovoltaic devices, said apparatus described in U.S. patent application Ser. No. 711,785 entitled Hybrid Semiconductor Processor And Gas Mixtures For Use In That And Other Processors, filed Mar. 14, 1985, now abandoned the disclosure of which is incorporated herein by reference. As disclosed in said application, microwave and/or radio frequency energy may be utilized to continuously deposit successive layers of semiconductor alloy material over very large area substrates.
Gas depletion is a significant problem encountered in the glow discharge plasma deposition of thick (over approximately 1000 angstroms) layers of semiconductor alloy material upon large area substrates. Gas depletion occurs under the influence of an electromagnetic field when certain gaseous components of the precursor process gas mixture decompose and deposit at greater rates than do other gaseous components of the precursor process gas mixture. Spatial irregularities or inhomogeneties in the composition of the deposition plasma and the resultant inhomogeneities and nonuniformities present in the deposited semiconductor alloy material can arise as a result of this type of gas depletion. These inhomogeneities and nonuniformities result in the deposition of layers of semiconductor alloy material having non-uniform electrical, chemical and optical properties. The problem of gas depletion is particularly manifested when a relatively thick (over 3000 angstrom) layer of narrow band gap silicon: germanium alloy material is being deposited from silane and germane precursor gases in a roll-to-roll processor. This is because the germane gas "cracks" much more easily than does the silane gas and the germanium is deposited therefrom at a much faster rate than is the silicon. Since it is necessary to deposit such a thick layer, the cathode must be very long (approximately seven feet) and the gaseous components which make up the precursor mixture must be kept uniform throughout that length. It is also to be noted that the problem of gas depletion is particularly significant in a microwave energized plasma deposition system insofar as the high energy of the microwave power results in high deposition rates in which the precursor components are even more readily cracked. Therefore, the use of microwave energy can be seen to exacerbate gas depletion problems with precursor components which do not crack at equal rates.
By way of a more specific illustration, a silicon: germanium alloy comprising approximately 60 percent silicon and 40 percent germanium has a band gap of approximately 1.4 eV and has significant utilization as the bottom layer of intrinsic semiconductor alloy material in a triple tandem photovoltaic device. Applicants have found that the use of a precursor gaseous mixture, which includes the typical process gases (i.e. silane and germane), presents significant problems in the preparation of such a 1.4 eV band gap silicon: germanium alloy. First of all, the difficulty in achieving the appropriate mixture of discrete precursor gases must be appreciated. In order to deposit a semiconductor alloy composition having approximately 60 percent silicon and 40 percent germanium, a gas mixture of silane and germane having far less than 40 percent germane must be employed since (as previously stated) the germanium from the germane gas deposits at a higher rate than does the silicon from the silane gas. Furthermore, the particular composition of this gas mixture is also dependent upon, inter alia, the deposition power and the geometry of the deposition system. A trial and error approach must therefore be employed to achieve the proper gas mixture for the deposition of a preselected semiconductor alloy composition which can be uniformly maintained for a prolonged period of deposition. In addition to the problems in formulating the initial mixtures of process gases, the rapid depletion of the germanium-containing precursor gas leads to spatial inhomogeneities and nonuniformities in the deposited film of semiconductor alloy material.
Various approaches to solving the problem of depletion of the germanium-containing gas relative to the silicon-containing gas have been implemented with varying degrees of success. For example, U.S. patent application Ser. No. 664,453 filed Oct. 24, 1984, entitled Cathode Assembly With Localized Profiling Capabilities, which application is assigned to the assignee of the instant invention and the disclosure of which is incorporated herein by reference, discloses an electrode for a glow discharge deposition system. The electrode is specifically designed with the capability of introducing any combination of precursor process gases at differing locations along the length thereof to substantially reduce the effects of the gas depletion. More particularly, the electrode functions to refresh and supplement discrete components of the process gas mixture at preselected longitudinal locations in the plasma region and thereby substantially obviates the effects of gas depletion. While satisfactory results are obtained with this system, (1) it is a "hardware solution", requiring modification of presently utilized deposition systems, and (2) it is not readily adaptable for microwave energized deposition systems insofar as the gas introduction is accomplished through the energized cathode plate itself, and most microwave deposition systems do not include an energized cathode, but rely upon the use of antennae or waveguides for the introduction of the microwave energy. Accordingly, it is desirable that the problem of gas depletion be addressed and solved for both microwave and radio frequency energized systems without the necessity of including extraneous hardware such as the cathode assembly of the aforementioned application.
In accordance with one aspect of the instant invention, it has been discovered that disilane, as well as other higher order silanes such as Si.sub.3 H.sub.8 Si.sub.4 H.sub.10 etc. and "substituted" higher order silanes such as fluorinated higher order silanes, collectively refered to herein as polysilanes, may be advantageously employed to eliminate problems of gas depletion. It has been found that polysilanes, under the influence of the electromagnetic field developed in a glow discharge deposition system, decompose and deposit the silicon therefrom at a higher rate than does silane or silicon tetrafluoride. Applicants have formulated a multi-component process gas mixture which includes a germanium-containing gas such as germane and a polysilane gas such as disilane. Under the influence of a glow discharge deposition plasma the germanium-containing gas and the silicon-containing gas decompose at a approximately similar rates to deposit a silicon: germanium alloy. Since the two gases decompose at approximately the same rate, the composition of the gas mixture will be essentially unchanged throughout the deposition process. Therefore, problems resulting from the differential depletion of the two components of the precursor gaseous mixture will be eliminated and the deposited alloy will be homogeneous and uniform in composition and exhibited properties. Additionally, since the component gases of the mixture decompose at approximately similar rates, the composition of the initial gas mixture will be reflected in the homogeneous and uniform stoichiometry of the alloy deposited across the surface of a large area substrate. It should thus be apparent that the use of the gas mixtures of the instant invention, which will be described in greater detail hereinbelow, confers signigicant advantages in the preparation of multi-component semiconductor alloys insofar as such mixtures eliminate the problems of gas depletion.
While it has been found that polysilanes, particularly disilane, may be utilized in conjunction with germanium-containing gases to eliminate problems of differential gas depletion, Applicants have also found that such polysilanes may be advantageously employed either with or without a germanium-containing precursor gas to produce a semiconductor alloy exhibiting high quality electronic properties. The improvements are manifested in (1) improved photovoltaic properties of the bulk material, such as high photoconductivity, a low density of states in the band gap and so forth, and (2) improved performance in photovoltaic devices in which the semiconductor alloys are incorporated, i.e., increased efficiency and decreased degradation.
Although the exact reason for the superior properties manifested by polysilane deposited materials is not completely understood, Applicants believe the reason is that polysilanes produce a greater number of desirable deposition species (species found in the plasma which provide high quality photovoltaic properties when deposited as semiconductor alloy material). It is known by those skilled in the art of plasma deposition that conditions which exist and the composition of species present within the plasma region are highly complex and uncontrollable. Some researchers in the field have described the plasma as being a "zoo" populated by numerous and rapidly changing exotic activated species of the precursor gaseous mixture introduced thereinto; the species formed by the fragmentation, ionization, radicalization and recombination of that gaseous mixture. It is known that the composition and characteristics of the deposited semiconductor alloy material will depend, inter alia, upon the particular excited species producing that deposit. For instance, certain species favor the deposition of tetrahedrally coordinated semiconductor alloy material, while other species favor the deposition of highly defective material, i.e., material having a significant number of defects, dangling bonds, strained bonds and/or vacancies therein. It is Applicants' belief that the polysilane gases of the instant invention produce a greater percentage of the desirable deposition species when exposed to plasma conditions, which species tend to deposit semiconductor alloy material having high quality photoelectrical properties. It is now a scientifically accepted principle that these high quality photoelectric properties are due, at least in part, to the presence of tetrahedral coordination in depositing material.
As disclosed in U.S. Pat. No. 4,226,898 of Ovshinsky, et al entitled Amorphous Semiconductors Equivalent To Crystalline Semiconductors Produced By A Glow Discharge Process, which patent is assigned to the assignee of the instant invention and the disclosure which is incorporated herein by reference, fluorine introduced into layers of amorphous silicon alloy semiconductor material operates to substantially reduce the density of localized defect states in the energy gap thereof and facilitates the addition of other alloying materials, such as germanium. As a result of the introduction of fluorine into the host matrix of the amorphous semiconductor alloy, the films so produced exhibit a number of favorable attributes similar to those of crystalline semiconductor materials. A fluorinated amorphous semiconductor alloy can thereby provide relatively high photoconductivity, increased charge carrier mobility, increased diffusion length of charge carriers, low dark intrinsic electric conductivity, and where desired, such alloys can be modified to help shift the Fermi level to provide substantially n or p type extrinsic electrical conductivity. As disclosed therein fluorinated amorphous semiconductor alloy materials can act like crystalline materials and be useful in the manufacture of photoresponsive devices such as, solar cells and current controlling devices including diodes, transistors and the like.
Fluorine is a "super-halogen", exhibiting the highest electronegativity of all of the elements, and is quantitatively and qualitatively different from the other halogens. As previously recognized and reported, fluorine acts in several ways to produce a better alloy material. For instance, fluorine can act in the bulk of a semiconductor alloy material to organize the local order of atoms to assure the formation of proper electronic configurations. Additionally, fluorine can passivate dangling bonds, reduce deviant bonding configurations and generally lower the density of states of a semiconductor alloy material into which it is incorporated.
Applicants have found that fluorine, in addition to its effect on the bulk of a semiconductor alloy material, can also act at the depositing surface of the semiconductor alloy material to improve the electronic properties of that material. The addition of a small amount of fluorine into the deposition atmosphere, or alternatively, the post deposition treatment of a semiconductor alloy material with fluorine can serve to reduce the number of surface states in that material. In this manner, a relatively small amount of fluorine can significantly improve the electronic, chemical and optical properties of a depositing layer of semiconductor alloy material.
It has now been recognized by Applicants that even small amounts of fluorine gas significantly affect the composition of the glow discharge plasma (separate from any influence the fluorine exerts in or on the depositing film), which plasma effects result in a significant improvement in the photoelectronic properties of the deposited semiconductor alloy material. More particularly, it has been found that the addition of even relatively small amounts of fluorine to the energetic plasma environment results in the deposition of semiconductor alloy material having significantly increased photoconductivity and a lower density of states in the band gap. Such semiconductor alloy material can be utilized to produce photovoltaic devices having high efficiency and long term stability.
The manner in which fluorine acts within the electromagnetic field established by a glow discharge is as complex as the plasma itself. As discussed previously, the plasma includes a multitude of highly excited species, some of which result in the deposition of high quality semiconductor alloy material whereas others of which result in the deposition of low grade semiconductor alloy material. It is now believed by Applicants that fluorine acts to organize the species in the plasma so as to produce a greater number of the optimum deposition species than are ordinarily present. Thus, fluorine may be said to act as a "plasma catalyst" facilitating, through its high electronegativity and high reactivity, the formation and maintenance of desirable free radical and/or ionic species in the plasma, which desirable species are then deposited as high quality semiconductor alloy material.
It is also hypothesized that fluorine acts in a catalytic manner upon the depositing surface of the layer of semiconductor alloy material to improve the photoresponsive properties of that layer by (1) insuring proper tetrahedral incorporation of the depositing atoms, (2) removing depositing species having undesirable or weak bonding configurations or (3) catalyzing the reformation of improper bonds. For example, fluorine may "etch away" certain species from the surface of the depositing material and recycle those species back into the plasma for re-excitation into the desirable tetrahedral configuration. Fluorine may thus be seen to act as a catalyst upon the surface of the depositing film, i.e., assuring proper bonding configurations thereof without being significantly incorporated in that film. Though the exact modus operandi of the fluorine atoms and molecules in the plasma chemistry is still subject to conjecture, it will be shown hereinbelow that the presence of relatively small amounts of fluorine in the highly energetic confines of the deposition plasma is extremely effective in conferring significant improvements in the layer of semiconductor alloy material deposited therefrom. The belief that fluorine plays a significant role in organizing plasma chemistry may be ascertained from the fact that improvements in the photoresponsive properties of semiconductor alloy material can be found when only less than one tenth of one percent of fluorine is incorporated into the matrix of that material. It is therefore one aspect of the instant invention to provide deposition plasma with a source of fluorine gas for organizing that plasma into desirable species and producing improved semiconductor alloy material.
The instant invention thus provides a precursor gaseous mixture adapted to provide for the deposition of high quality semiconductor alloys. The mixture includes a silicon-containing gas such as silane and a fluorine-containing gas such as silicon tetrafluoride, fluorosilanes (including fluoropolysilanes), fluorine, flurocarbons; and, in the case where doped alloy materials are being prepared, fluorinated dopant gases such as boron fluorides, phoshorous fluorides, aluminum fluorides and the like. The gas mixture may include a germanium-containing gas if a narrow band gap silicon:germanium alloy is being prepared. It is necessary to mention at this juncture that the narrow band gap silicon:germanium alloys generally exhibit poorer photovoltaic properties than do their silicon alloy counterparts. This is mainly due to the presence of a higher density of defect states in band gaps thereof, particularly deep states. Applicants, in formulating their inventive theories, speculated that fluorine, if properly introduced into the plasma with the properly selected precursor gaseous species of germanium, silicon and diluent gas, could act as a catalyst and promote tetrahedral coordination of the preferred deposition species so as to deposit a narrow band gap (about 1.45 eV) semiconductor alloy material having photovoltaic properties as good as the properties of the about 1.7 eV material. It is a notable feature of the instant invention that only relatively small amounts of the fluorine-containing gas need be employed to secure the aforementioned advantages therefrom.
In accordance with another aspect of the instant invention, it has been found that a precursor gaseous mixture which includes a polysilane compound, such as disilane, and a source of fluorine may be utilized to deposit a layer of semiconductor alloy material having improved photoelectric properties, thereby making such an alloy highly desirable for use in a photovoltaic cell. The polysilane gas and the fluorine-containing gas have thus been found to combine synergystically to still further improve the resultant semiconductor alloy material.
In keeping with the mechanics of the previously discussed model, it is postulated that the plasma decomposition of the polysilane compound produces a favorable number of desirable deposition species and the fluorine present in the plasma acts as a catalyst or organizer to (1) facilitate the creation of still more favorable deposition species, (2) facilitate the deposition of these desirable species in tetrahedral coordination and bonding configurations and (3) act upon the surface of the depositing alloy material to produce a high quality deposit. Additionally, as described hereinabove, the use of polysilanes results in a higher deposition rate, thereby improving the operational efficiencies of the deposition process and eliminating problems of gas depletion when silicon:germanium alloys are being prepared. It should be noted again that fluorine may also be directly incorporated into the polysilane molecule itself. That is to say, the polysilane precursor gas may be of the general formula: Si.sub.x R.sub.y H.sub.z where R is chosen from the group consisting essentially of hydrogen, fluorine, chlorine, iodine, bromine and combinations thereof, x is an interger greater than one, y and z are each an integer or zero, and y+z=2x+2. It should be noted that the foregoing formula is meant to include metastable compounds and species as well as stable molecules. Fluorinated polysilanes including metastable species thereof, as well as polysilanes produce a semiconductor alloy material characterized by increased photoconductivity and a decreased density of defect states in the band gaps thereof.
Note that Applicants are not claiming to be the first to utilize disilane gas in a vapor deposition system for the fabrication of semiconductor alloy material. As early as 1981, the use of disilane in a glow discharge deposition process has been reported in numerous papers and patents. For instance, Scott, et al, in a paper entitled "Deposition and Doping of a-Si:H from Si.sub.2 H.sub.6 Plasmas" published in Tetrahedrally Bonded Amorphous Semiconductors, pp. 6-9 (1981) reported upon the use of disilane for the preparation of silicon alloy films. However, Scott, et al discovered and were reporting on the fact that the use of disilane in such a manner results in a high deposition rate, which deposition rate is significant when depositing relatively thick layers of silicon alloy material. Although Scott, et al had noticed slight differences in the characteristics of the disilane prepared film, vis-a-vis, silane prepared films, they were primarily concerned with securing the high deposition rates which are necessary for depositing a 25 micron thick layer of semiconductor alloy material which is required for a copier drum. It should be noted that nowhere in the Scott, et al paper is there reported (1) the use of disilane in conjunction with a germanium-containing gas to produce a silicon:germanium alloy, (2) the problem of gas depletion which Applicants first discovered and report herein, (3) the use of fluorine as either a catalyst in the plasma or as a density of states reducer in the deposited silicon alloy film and (4) the fabrication of any type of photovoltaic device from the deposited material.
Similar results are reported in U.S. Pat. No. 4,363,828 of Brodsky, et al (Brodsky is one of the authors of the aforedescribed Scott, et al paper). The Brodsky, et al patent discloses the use of disilane in an inductively coupled radio frequency plasma deposition system in which a silicon alloy film is prepared. As in the Scott, et al paper, there is no disclosure of (1) the use of fluorine for any purpose whatsoever, (2) the use of disilane in conjunction with the germanium-containing gas to obviate the problem of gas depletion, (3) the preparation of improved photovoltaic devices by the use of disilane alone or in combination with fluorine and (4) the use of fluorine as either a catalyst in the plasma or as a density of states reducer in the deposited silicon alloy film.
The Canon Corporation reported upon the preparation of photoconductive drums for electrophotographic copiers in U.K. Patent Application Nos. 2,077,451A and 2,087,930A. These publications disclose the preparation of layers of photoconductive semiconductor alloy material from precursor gaseous mixtures of SiH.sub.4 SiF.sub.4 and a diluent energized in a radio frequency glow discharge process. It is also broadly suggested in these publications that disilane could be similarly employed to fabricate those layers of photoconductive material. However, no examples of a deposition involving disilane, and likewise no operational parameters of the xerographic drum thus prepared, are reported. It should be noted that because the Canon publications are solely oriented toward the preparation of layers of photoconductive amorphous silicon alloy material for use in electrophotographic drums, the material therein disclosed is unsuitable for the fabrication of effective photovoltaic devices insofar as it includes too great a density of defect states in the band gap thereof. It is also noteworthy that the researchers at Canon did not (1) employ fluorine in combination with the disilane as either a catalyst in the plasma or as a density of states reducing element in the deposited semiconductor alloy film (as is evidenced by the fact that they neither were not fabricating material for photovoltaic purposes nor was their material (due to the high density of defect states) capable of use in photovoltaic applications), (2) prepare silicon:germanium alloy materials and (3) report on the use of any polysilane components in the precursor gaseous mixture.
In yet another published U.K. patent application, Ser. No. 2,081,745A, researchers at Canon Corporation reported on another process for the preparation of amorphous silicon alloy materials. Disclosed in the publication is a d.c. or low frequency a.c. (i.e., several Hz) energized plasma deposition process. As disclosed, irradiation with the appropriate wavelength light is utilized to photoconductively discharge built up potential in the silicon alloy material, which potential limits the rate of deposition. Although the example given therein is limited to the use of silane, the disclosure also mentions the possibility of utilizing disilane or germane from which to fabricate amorphous semiconductor alloys. It should be noted that the disclosure is strictly oriented toward the production of a photoconductive copier drum and accordingly does not report on any material or device parameters which are directly applicable to photovoltaic technology. Furthermore, the disclosure (1) provides no examples of the use of disilane, (2) makes no mention of the use of fluorine as either a catalyst in the plasma or as a density of states reducing element in the deposited semiconductor alloy film and (3) does not discuss the fabrication of silicon:germanium alloy materials from non-depleting precursor gaseous mixtures. As with previously described references, disilane is discussed only with a view toward increasing the rate of deposition of the semiconductor alloy material because the photoconductive material of a copier drum must be deposited to a thickness of about 25 microns, as compared to the approximately one-half micron thickness which is deposited in the fabrication of photovoltaic cells.
In a paper entitled "Photoconductive a-SiGe:F:H and Transparent IrO.sub.x for High Efficiency Amorphous Solar Cells", published in the Technical Digest of the International PVSEC-1, Kobe, Japan (1984) pp. 429-432, Oda, et al have reported on the fabrication of a narrow band gap silicon:germanium:fluorine: hydrogen alloy for photovoltaic uses. The alloy is manufactured from a mixture of silicon tetrafluoride, germanium tetrafluoride and hydrogen. As stated by Oda, et al in the body of the paper, various combinations of gases were examined as precursors for the deposition of the alloy, said gases including silane, disilane, silicon tetrafluoride, germane and germanium tetrafluoride. However, Oda, et al reported that the best results were obtained through the use of silicon tetrafluoride, germanium tetrafluoride and hydrogen, where the concentration of hydrogen is typically 30%.
While Oda, et al have admittedly examined disilane containing mixtures for use in the fabrication of narrow band gap silicon:germanium alloys, they failed to recognize the advantages of disilane in terms of (1) the preparation of a significantly improved photovoltaic material, (2) the addition of fluorine to improve tetrahedral coordination in the deposited photovoltaic material and (3) obviating the gas depletion problem first recognized and described by Applicants in the instant application. While Oda, et al noted that the stoichiometry of the deposited semiconductor alloy film is dependent upon that of the precursor gaseous mixture, they did not recognize the problem of gas depletion, presumably because they were fabricating small area samples (under one square centimeter) in a nonproduction mode. For small area depositions, gas depletion problems will not be significant as compared to the gas depletion problems which Applicants encounter in the continuous production of semiconductor alloy material on an elongated web. It should also be abundantly clear, because of the absence of any discussion thereabout, that Oda, et al did not recognize fluorine's unique and synergistic utility as a plasma catalyst for organizing local order of the deposition species present in the highly energetic plasma and for insuring tetrahedral coordination of the depositing species.
In contrast to all the aforedescribed prior art, the instant invention provides a unique class of precursor gaseous mixtures which are particularly adapted for fabricating photovoltaic cells from the glow discharge deposition of successive layers of semiconductor alloy material. When (1) disilane is utilized as the silicon precursor gas, (2) fluorine may be added as a catalyst or organizer in the plasma, as well as a density of states reducer in the depositing film, and (3) hydrogen gas may be used as a diluent in order to provide precursor gas mixtures from which particularly efficient photovoltaic cells may be fabricated. Through the use of such precursor gaseous mixtures, problems of differential process gas depletion, which problems typically occur in the preparation of narrow band gap silicon:germanium alloys, are substantially eliminated. And finally, through the use of such precursor gaseous mixtures, the rate at which high quality semiconductor alloy material may be deposited is substantially increased.
Applicants have stated, explicity as well as impliedly, throughout the course of the preceeding discussion, that the prior art semiconductor alloy material (fabricated through the fluorine-free glow discharge deposition of a disilane-containing precursor mixture pursuant to the processes described by the aforementioned references) is not of the quality necessary to fabricate a high efficiency, high stability photovoltaic device. Because the reasons Applicants claim to have developed superior films may not be readily apparent in light of the confusing number of prior art references and the preceding discussion of that prior art, a more detailed explanation follows.
The explanation is best accomplished with specific reference to the Oda, et al publication, which publication describes an amorphous silicon:germanium:fluorine:hydrogen alloy allegedly having a photoconductivity of 9.times.10.sup.-5 inverse ohms-centimeters. Applicants state that the narrow band gap semiconductor alloy material described in the instant application, although having a photoconductivity of only approximately 3.times.10.sup.-5 inverse ohms-centimeters, is a much better photovoltaic material than that of Oda, et al.
Applicants' material is superior to Oda, et al's material in spite of its slightly lower photoconductivity because of the fact that photoconductivity is not, in and of itself, a sufficient measure of the photovoltaic quality of deposited semiconductor alloy material. First of all, photoconductivity data does not give any indication of minority carrier lifetime, an important parameter in characterizing the quality of material adapted for use in a photovoltaic cell. Secondly, photoconductivity is dependent upon the position of the Fermi level (and hence the activation energy) of the particular semiconductor alloy material under consideration. More specifically, as the activation energy decreases, the photoconductivity will increase because charge carriers are moved closer to the conduction band. Therefore, the mere existence of high photoconductivity in a material does not necessarily mean that the material will make a high efficiency photovoltaic device.
As reported in another paper of Oda, et al entitled "Hole Transport in Glow Discharge a-SiFe.sub.x H(F) Alloys", the narrow gap material discussed above was characterized as having a high density of states adjacent the conduction band, as is indicated by the high-sub band gap absorption of the semiconductor alloy material prepared by Oda, et al. This high density of defect states reduces the lifetime of minority charge carriers and increases the recombination rate of charge carriers within the band gap. Simply stated, the photoconductivity of Oda, et al's material is relatively high because the Fermi level has been moved closer to the conduction band, thereby reducing the activation energy necessary to promote charge carriers thereto; however, photovoltaic quality is low because of the presence of a high density of defect states.
Therefore, Applicants' narrow band gap semiconductor alloy material, which material has a photoconductivity of approximately 3.times.10.sup.-5 inverse ohms-centimeters while also having an activation energy of about 0.65-0.75 eV (the Fermi level is pinned proximate mid-gap) and a low density of defect states, is superior to Oda, et al's semiconductor alloy material even though the photoconductivity of Oda, et al's material is artifically better than Applicants'. The simple fact is that Oda, et al's semiconductor alloy material has an increased density of defect states adjacent the conduction band edge which severely reduces minority carrier lifetime. In contrast thereto, Applicants' low band gap (1.45 eV) semiconductor alloy material is of the same high quality (with respect to the number of defect states in the band gap thereof and the photoconductivity thereof as the best 1.75 eV material heretofore produced (and far superior to any low band gap semiconductor alloy material heretofore produced).
The aforedescribed advantages of the instant invention as well as many other advantages will be made clear by the Detailed Description of the Drawings and the claims which follow.